Levitt's CO breath test in the differential diagnosis of chronic isolated hyperbilirubinemia.

A key component of the differential diagnosis of isolated hyperbilirubinemia (HB) is distinguishing between hemolytic and non-hemolytic types. Routine hemolysis screening markers have unsatisfactory sensitivity and specificity. Erythrocyte (RBC) lifespan shortening, the gold standard marker of hemolysis, is seldomly measured due to the cumbersome and protracted nature of standard methods. A new Levitt's CO breath test method may enable simple, rapid RBC lifespan measurement. In this pilot prospective diagnostic study, Levitt's CO breath test was evaluated to discriminate hemolytic from non-hemolytic HB in adults. One hundred and thirty eligible non-smoking adult patients who were aged 18 or older, referred for chronic (>6 months) isolated HB or had a known diagnosis of isolated HB of a rare cause, were recruited, including 77 with non-hemolytic HB and 53 with hemolytic HB. ROC curve analysis was applied to determine the optimal cutoff for discriminating between hemolytic and non-hemolytic HB, and the performance was calculated. Results showed that the mean RBC lifespan in non-hemolytic HB (93 ± 26 d) was reduced (= 0.001 vs. normal reference value of 126 d), but longer than that in hemolytic HB (36 ± 17 d;= 0.001). RBC lifespans did not differ significantly between 26 patients with simple hemolytic HB (32 ± 14 d) and 27 patients with a Gilbert syndrome comorbidity (40 ± 18 d). ROC curve analysis revealed an optimal lifespan cutoff for discriminating between hemolytic and non-hemolytic HB of 60 d (AUC = 0.982), with a diagnostic accuracy of 95.4%, 94.3% sensitivity and 96.1% specificity respectively. These results indicate that Levitt's CO breath test seems to be very sensitive and specific for detecting hemolysis in adult patients with chronic isolated HB, and could enable simple, rapid, and reliable differential diagnosis of isolated HB. A large-scale validation study of the method is warranted.